Freemium optimization skill

Freemium Optimization

Overview

"Free" gets users. "Free" doesn't pay the bills. The gap between free and paid is where most PLG companies die — and the common mistake is optimizing conversion rate in isolation rather than full-funnel math. Kyle Poyar's 2026 Free-to-Paid Conversion Report (200 B2B products, with ChartMogul and ProductLed) shows that freemium drives roughly 90 signups per 1,000 visits vs 45 for free trials, meaning a lower per-signup conversion rate (8-12% GREAT for freemium vs 10-15% for no-CC trial) can still produce more customers end-to-end. Optimizing conversion rate in isolation causes teams to abandon freemium prematurely.

The second mistake: skipping activation instrumentation and jumping straight to conversion optimization. ProductLed's State of B2B SaaS 2025 (446 companies) found only 34% of companies track activation — but you cannot improve conversion from an experience users have not completed. OpenView's 2023 Product Benchmarks (with Pendo) found that tracking PQLs/PQAs increased the likelihood of fast growth by 61% — the single most influential lever in their study.

When to Use

• "Optimize freemium conversion"
• "Design free-to-paid upgrade path"
• "Set up PQL scoring"
• "Improve activation flow"
• "Choose between freemium and free trial"
• "Reduce time-to-value"
• Triggers on: "freemium optimization", "free trial conversion", "PQL scoring", "activation flow", "paywall design", "freemium to paid"

Authoritative Foundations

• Kyle Poyar (Growth Unhinged) + ChartMogul + ProductLed — 2026 Free-to-Paid Conversion Report (200 B2B products). Primary benchmark source for this skill. Provides median and GOOD/GREAT conversion rates by model type and credit-card-gate status, and the full-funnel signup-rate data that makes model choice a visits-to-customers calculation rather than a conversion-rate comparison. Key finding: 57% of products lead with free trial vs 26% freemium. See references/framework-notes.md for the full benchmark table by model and ACV bracket.
• ProductLed — State of B2B SaaS 2025 (446 companies). Source for PQL adoption and activation tracking data. Only 24-25% of PLG companies use PQLs, but PQL users see roughly 3× higher free-to-paid conversion. Only 34% track activation. Companies with "highly intentional" free models (8+/10) report 57% better free-to-paid conversion than unintentional ones (3 or below). Intentional design means the free tier showcases core value, creates natural upgrade paths, has deliberate value limits, and makes upgrade benefits visible inside the free experience.
• OpenView Partners — 2023 Product Benchmarks (with Pendo, ~1,000 participants). Tracking PQLs/PQAs increased likelihood of fast growth by 61% — the single most influential lever measured. Outreach to free signups adds 28%; a dedicated growth team adds 17%; over-relying on paid acquisition is inversely correlated with fast growth.
• Wes Bush — Product-Led Growth / Product-Led Onboarding. Provides the bowling-alley onboarding model — guide rails (in-app prompts) and bumpers (email nudges) keep users on the path to their activation moment. The principle that the free tier must deliver the core value experience, and that time-to-value reduction is the primary activation lever. Used in Phase 2 to design the activation flow.
• Dharmesh Shah (HubSpot) — Freemium as Flywheel Attract. Free CRM/tier as top-of-flywheel acquisition; upgrade on seats, automation, integrations. Pair with inbound-triage for PQL→SQL handoff. references/dharmesh-shah-hubspot-inbound.md.

Step-by-Step Process

Phase 1: Model Decision with Full-Funnel Math

Before optimizing, select (or audit) the right model using full-funnel math — not conversion rate alone.

Model	Signups / 1,000 visits	GOOD conv rate	GREAT conv rate	When to use
Freemium (forever free)	~90	3-5%	8-12%	Core value deliverable in free; large TAM; viral/network effects
Free trial (no CC)	~45	4-6%	10-15%	Complex product; hands-on trial needed; mid-market motion
Free trial (CC required)	~15-20	25-35%	50-60%	High-intent buyers; self-serve checkout; lower volume acceptable

Source: Poyar/ChartMogul/ProductLed 2026. The CC-gated model shows the highest conversion rate but suppresses signup volume 4-6×. Run the full-funnel math (signups × conversion rate = customers per 1,000 visits) before choosing it. See references/framework-notes.md for the benchmark table by ACV bracket.

Model intentionality test (ProductLed 2025): Score your free model 1-10 on intentionality. Companies scoring 8+ report 57% better free-to-paid conversion than those scoring 3 or below. The test: does the free tier showcase core value? Are value limits deliberate? Are upgrade benefits visible inside the free experience?

Phase 2: Activation Design

Define the activation moment before measuring anything else. Activation is the point where a free user experiences your product's core value for the first time. Without a named definition, there is no baseline and no conversion target.

1. Name the activation event as a specific product event — for example: "User completes first export," "Two team integrations connected," "Live-data report viewed."
2. Measure current time-to-activation (median and p90). Target: median under 7 days.
3. Apply Wes Bush's bowling-alley model: in-app guide rails for the critical path, email bumpers for users who stall between steps.
4. Fill empty states with sample data showing what the user will see at activation — empty states are the leading cause of early drop-off.
5. Reduce steps-to-activation: every screen added between signup and the activation moment reduces completion rate.

Track activation before running conversion experiments. Only 34% of PLG companies currently instrument activation (ProductLed 2025). If you are not in that 34%, instrument first.

Phase 3: Paywall Design and Placement

Paywall timing: After activation, before habit. A user who has experienced core value but is not yet a daily user is at peak upgrade motivation. Show the paywall before activation and they leave; after habit is formed, urgency diminishes.

Paywall triggers — map to product events, not just calendar time:

• Usage limit reached (volume-based gate)
• Feature in paid tier requested (feature gate)
• Team invite sent (collaboration requires paid tier)
• Time-based trial expiration (fallback trigger only)
• Export/share/publish (output-based gate)

Free tier design principle: Enough to love, not enough to stay forever. Free must deliver the core value experience so users activate and form intent to buy, but leave a natural job undone that the paid tier completes.

Phase 4: PQL Scoring and Routing

OpenView 2023: tracking PQLs/PQAs increased likelihood of fast growth by 61% — the highest-impact lever in the benchmark. Despite this, only 24-25% of PLG companies currently use PQLs (ProductLed 2025).

Default scoring skeleton — calibrate weights to your product:

Signal	Weight
Activation event completed	30
Usage depth (key feature engagement)	25
Usage frequency (days active / last 14 days)	20
ICP fit (firmographic match)	15
Expansion signals (team invites, integrations connected)	10

Routing thresholds:

• Score ≥ 70: immediate sales or high-touch outreach (PQL)
• Score 50–69: automated nurture sequence + SDR monitoring
• Score < 50: focus on activation, not conversion

A PQL model that scores users but does not trigger a defined action within an SLA delivers no value. Define routing and response SLAs before launch. See references/framework-notes.md for PQL trigger examples by product type.

Phase 5: Experiment Backlog

Structure experiments against full-funnel stages, ordered from top of funnel to bottom:

1. Activation experiments: Reduce time-to-activation; test guided vs self-directed onboarding; test sample-data pre-fill; test step reduction.
2. Paywall experiments: Test trigger placement (usage limit vs feature gate vs time); test paywall copy and upgrade-value framing; test pricing page layout.
3. PQL routing experiments: Test outreach timing (immediate vs 24-hour delay post-score); test channel (email vs in-app vs sales call).
4. Model experiments: If evidence supports it, test CC-gate vs no-CC gate on a traffic split — measure customers-per-1,000-visits as the primary metric, not conversion rate.

Output Format

Freemium optimization plan containing: model decision with full-funnel math (signups × conversion rate = customers per 1,000 visits) benchmarked against Poyar/ChartMogul data; activation definition with named product event, instrumentation plan, and time-to-activation baseline; paywall trigger map with product-event-to-gate assignments and timing rationale; PQL scoring model with signal weights, routing thresholds, and response SLAs; benchmark-anchored experiment backlog ordered by funnel stage (activation first, then conversion, then routing).

Quality Check

Before delivering, verify:

• Model choice is justified with full-funnel math (visits → signups → conversions → customers), not conversion rate alone
• Activation event is named as a specific product event, not "user engages with the product"
• Paywall placement is after the defined activation event, not at signup
• PQL model has at least 5 signal dimensions with documented weights and routing thresholds
• Every benchmark cited names the model type (freemium vs trial vs CC-gated) and the source report
• Experiment backlog is ordered by funnel stage with activation experiments first

Common Pitfalls

1. Optimizing conversion rate instead of full-funnel customers-per-visit. A CC-gated trial shows 50%+ conversion but suppresses signup volume 4-6×; end-to-end customer count can be lower than freemium. Fix: always run the full-funnel math before declaring a model "better."
2. Skipping activation instrumentation. You cannot improve conversion from an experience users have not completed. Fix: define and instrument the activation event before running any conversion experiment — only 34% of PLG companies currently do this (ProductLed 2025).
3. Free tier that replaces paid. If free delivers 90% of what paid delivers, users do not upgrade. Fix: apply the intentionality test — free showcases core value, paid completes the job; deliberate value limits and visible upgrade benefits inside the free experience are required.
4. PQL scoring without routing. A model that scores users but triggers no action within a defined time window delivers no revenue lift. Fix: define routing thresholds and response SLAs before launch; OpenView confirms outreach to free signups adds 28% to fast-growth likelihood.

Execution Artifacts

• references/framework-notes.md — Conversion benchmarks, PQL triggers, activation template
• templates/output-template.md — Deliverable shell for agent output
• scripts/check-output.py — Lightweight deliverable validator Lifecycle (Acquisition → Activation): references/activation-playbook.md · references/gtm-lifecycle-stages.md · Pattern 18 in using-gtm-skills Canonical lifecycle: references/gtm-lifecycle-stages.md (Acquisition, Activation) · references/activation-playbook.md · references/lifecycle-metrics-by-stage.md

Related Skills

• plg-strategy, growth-experimentation, onboarding-flow, pricing-psychology, a-b-testing